VL

vllm-d3

GGUFGGUF by ggml-orgUpdated 2026-10-26
Download on Hugging Face
Parameters26.9B
Context256K
Layers64
Architectureqwen35
Smallest quant19.4 GB
Quick answer

vllm-d3 needs about 22.1 GB of memory at Q4_K_M (a 19.4 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 3 quants, from 19.4 GB (Q4_K_M) to 53.8 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 8.6 GB.

vllm-d3: 27B multimodal decision model with text, image, and video input

vllm-d3 is a 26.9B parameter Qwen‑35 architecture model. It can answer multiple-choice questions, yes/no, or ratings in a single call. It supports text, images, and videos, returning probabilities without generating text.

Good at

  • Multimodal decision making
  • Single‑call multi‑question responses
  • Probability‑based answers

Limits

  • No text generation
  • Requires large VRAM for 8k context
  • No explicit language support list

Best for

  • Decision support systems
  • Rapid multi‑modal QA
  • Probabilistic inference tasks

Who it’s for. Developers building decision‑oriented applications that need to process text, images, and videos together. The model is ideal for systems that require probability outputs rather than generated text.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M19.4 GB 22.1 GB
Too big Get file →
Q8_028.6 GB 31.2 GB
Too big Get file →
BF1653.8 GB 56.5 GB
Too big Get file →

For 8 GB VRAM use Q4_K_M (19.4 GB file, ~22.1 GB memory at 8k context). For 12 GB VRAM use Q8_0 (28.6 GB file, ~31.2 GB memory). BF16 needs ~53.81 GB file and ~56.5 GB memory, suitable only for high‑end GPUs.

What GPU do you need for vllm-d3?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: no quant fits at an 8K context.
  • 24 GB VRAM: the best quant that fits is Q4_K_M (19.4 GB file, about 22.1 GB in use).
  • 32 GB VRAM: the best quant that fits is Q8_0 (28.6 GB file, about 31.2 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (28.6 GB file, about 31.2 GB in use).

Which vllm-d3 quant should you choose?

Q4_K_M is the usual starting point for vllm-d3: it balances file size and quality well. Here it is a 19.4 GB file.

The Q8_0 file is 28.6 GB, about 1.5 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen35
Parameters26.9B
Layers64
Trained context262,144 tokens
KV cache (fp16)256 KB per token

Source

Quantized byggml-org
Base modelvllm-sr/d3
Licenseapache-2.0
Downloads (HF)25,434
Last updated2026-10-26

Frequently asked questions about vllm-d3

How much VRAM does vllm-d3 need?

About 22.1 GB at Q4_K_M with an 8K context window. That is the 19.4 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run vllm-d3 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 22.1 GB, so you would need a larger GPU or CPU offloading.

What is the file size of vllm-d3 GGUF?

It depends on the quant: Q4_K_M is 19.4 GB, Q8_0 is 28.6 GB, BF16 is 53.8 GB.

What context length does vllm-d3 support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 2.1 GB at 8K and 8.6 GB at 32K.

Where can I download vllm-d3 GGUF?

The files are on Hugging Face in the ggml-org/vllm-d3-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does vllm-d3 use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run vllm-d3 on 8 GB VRAM?

Use the Q4_K_M quantized file; it requires about 22.1 GB memory at 8k context, so 8 GB VRAM is insufficient.

Does vllm-d3 support text generation?

No, it returns probabilities for answers without generating text.

Other GGUFs of d3

Similar models

Browse all →

More from ggml-org

All by ggml-org →

Other medium models

All medium models →